For cross-border eCommerce businesses looking to do business within the European Union, keeping up with the complex and constantly evolving world of trade regulations — from codes, declarations, duty and tax calculations to customs clearance requirements that may vary from one country to another — can be a daunting task.
The EU Value Added Tax (VAT) rule, under which imports valued at €22 euros were not subject to the tax, was removed in July last year, further complicating the regulatory environment for many eCommerce firms that must now endure heavy bureaucracy to pay VAT on all their imports.
“Most of eCommerce goods that were declared were under €22 [and therefore exempt from tax],” Egon Veermae, chief operating officer at Estonia-based eCommerce compliance firm Eurora Solutions, told PYMNTS in an interview.
He added that these stringent regulations are in stark contrast to the U.S., where firms benefit from a $800 threshold before they are subject to tax rules.
Another challenge EU cross-border firms are facing is the requirement to now declare all the goods to customs. According to Veermae, this leads to the complex task of having to calculate different VAT rates based on where the consignee is based — for example, the standard VAT rate is 19% in Germany, while in Estonia, it’s 20% — in addition to other tax requirements firms might be subject to.
At the end of the day, companies struggle to keep up, leaving them in dire need of a solution that helps them to calculate their taxes correctly. It’s a solution that the artificial intelligence (AI) and machine learning (ML)-backed firm is offering to clients through its proprietary platform that fully automates tax, compliance and customs services.
“We have integrations with the different European customs authorities because there is not a [centralized] customs authority in Europe,” Veermae explained. “Beyond these technical integrations, there’s also the language barrier we help tackle.”
Unlike some competitors, Eurora Solutions’ automated solution is accompanied by a customer support service to help customers to navigate new regulations.
“When we onboard the customer, we always appoint a special tax accountant to you who can always answer any questions you might have,” he noted.
Middle East Expansion
Last month, the cross-border eCommerce compliance platform snagged $40 million in Series A funding in an oversubscribed round that was considered one of the largest financing rounds of its kind in the Estonian startup ecosystem.
Read more: Eurora Solutions Snags $40M for eCommerce Compliance Platform
According to Veermae, parts of the funds have been earmarked for expansion into Middle Eastern markets, which hold huge potential for business growth but require a good understanding of the local culture and how to successfully build a business in the region.
“We want to find our niche and identify how we can help businesses,” he said. “That is the biggest challenge, in addition to ensuring a good penetration in these markets.”
However, another challenge will come from regulators and the pressure governments are putting on cross-border businesses to collect more taxes.
“Countries now require eCommerce platforms to start collecting taxes cross-border, I think this is going to be the biggest challenge,” Veermae said. “This was implemented in the U.K., now in Europe, and next year Singapore will start to use the same logic.”
Sign up here for daily updates on all of PYMNTS’ Europe, Middle East and Africa (EMEA) coverage.
The AI world is hitting our wallets with prices that might make even Wall Street veterans do a double take. While banks have long charged hefty fees for premium services, artificial intelligence (AI) companies are now joining the high-price club with some truly jaw-dropping figures. What happened to all that talk about AI making things cheaper and more efficient?
OpenAI, the company behind ChatGPT, is planning to launch AI “agents” with price tags that might make you spill your coffee. According to recent reports, it’s eyeing a tiered pricing plan that starts at $2,000 monthly for basic agents aimed at “high-income knowledge workers.” Need something a bit more sophisticated for software development? That’ll be $10,000 per month. And if you want the top-tier, PhD-level research agent, prepare to shell out an astonishing $20,000 every month.
To put that in perspective, OpenAI’s current premium offering, ChatGPT Pro, costs $200 monthly and is reportedly still losing money. The jump from $200 to $20,000 is a 100-fold increase. OpenAI CEO Sam Altman has apparently acknowledged they need to “charge much more than $200 a month” for these advanced agents. The company is banking on these agent products to generate between 20% and 25% of its long-term revenue.
SoftBank seems convinced these prices make sense — the investment company has already committed to spend $3 billion on OpenAI’s agents this year alone. The justification? These AI assistants supposedly do work comparable to highly paid professionals.
The pricing inconsistency across the AI landscape is enough to give you whiplash. Some companies are stuffing AI features into existing products and bumping prices, while others charge only when the AI completes a task.
For comparison, a coding agent from startup Cognition called Devin costs about $500 monthly — significantly less than OpenAI’s planned $10,000 offering for similar functionality. This kind of price variation leaves customers scratching their heads about what’s reasonable.
Whatever approach companies take, one thing is clear: AI doesn’t come cheap. The massive data centers powering these systems cost between $500,000 and over $1 billion annually, according to McKinsey & Company. Those specialized AI chips? They run between $10,000 and $30,000 each. A single server rack packed with these chips can easily cost over $500,000 before even turning on the power.
The million-dollar question (sometimes literally): Are these AI services actually worth these premium prices? Companies claim these tools can replace work done by highly paid professionals. A Ph.D.-level research agent at $20,000 monthly costs about the same as hiring a human researcher with advanced credentials.
Defenders of these prices point out that AI assistants work 24/7 without breaks or benefits. They don’t quit unexpectedly or need training when you change projects. They can potentially process information faster than any human researcher.
However, skeptics wonder if these theoretical advantages really justify the sky-high costs. Can an AI agent perform complex tasks requiring deep understanding and original thinking? With their regulatory requirements, financial institutions need to be especially careful about relying too heavily on systems that sometimes generate incorrect information.
As artificial intelligence transforms from a cool experiment to an essential business tool, we’re witnessing a gold rush mentality in pricing. While there are legitimate costs behind developing and running these systems, the current prices seem to include a healthy dose of “what the market will bear” thinking.
Considering that, according to PYMNTS Intelligence data, the same market is starting to report positive returns on investments (ROI) on its GenAI investments, it might be willing to bear the higher costs.
Will these prices eventually come down as competition increases and technology improves? History suggests they might. But for now, if you want the cutting edge of AI, be prepared for some serious sticker shock. The smartest approach might be testing these premium-priced tools selectively while keeping a healthy skepticism about whether that shiny new AI assistant is really worth its weight in digital gold.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.